[PATCH] D106447: [DAGCombine] DAGTypeLegalizer::GenWidenVectorLoads(): make use of dereferenceability knowledge

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/copy-illegal-type.ll:165
 ; SI:       ; %bb.0:
-; SI-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x11
+; SI-NEXT:    s_load_dwordx8 s[4:11], s[0:1], 0x11
+; SI-NEXT:    s_waitcnt lgkmcnt(0)
----------------
This looks alarming. It's loading 32 bytes instead of 8.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D106447/new/

https://reviews.llvm.org/D106447